Structure and Function of Quinones in Biological Solar Energy

Oct 16, 2009 - Differential Pulse Voltammetry, EPR, and Hyperfine Sublevel Correlation (HYSCORE) ... function in energy transduction remain unclear...
0 downloads 0 Views 1MB Size
J. Phys. Chem. B 2009, 113, 15409–15418

15409

Structure and Function of Quinones in Biological Solar Energy Transduction: A Differential Pulse Voltammetry, EPR, and Hyperfine Sublevel Correlation (HYSCORE) Spectroscopy Study of Model Benzoquinones Amanda M. Weyers, Ruchira Chatterjee, Sergey Milikisiyants, and K. V. Lakshmi* Department of Chemistry and Chemical Biology and The Baruch ’60 Center for Biochemical Solar Energy Research, Rensselaer Polytechnic Institute, Troy, New York 12180 ReceiVed: July 31, 2009; ReVised Manuscript ReceiVed: September 16, 2009

Quinones are widely used electron transport cofactors in photosynthetic reaction centers. Previous studies have suggested that the structure of the quinone cofactors and the protein interactions or “smart” matrix effects from the surrounding environment govern the redox potential and hence the function of quinones in photosynthesis. In the present study, a series of 1,4-benzoquinone models are examined via differential pulse voltammetry to provide relative redox potentials. In parallel, CW and pulsed EPR methods are used to directly determine the electronic properties of each benzoquinone in aprotic and protic environments. The shifts in the redox potential of the quinones are found to be dependent on the nature of the substituent group and the number of substituent groups on the quinone molecule. Further, we establish a direct correlation between the nature of the substituent group and the change in electronic properties of the benzosemiquinone by analysis of the isotropic and anisotropic components of the electron-nuclear hyperfine interactions observed by CW and pulsed EPR studies, respectively. Examination of an extensive library of model quinones in both aprotic and protic solvents indicates that hydrogen-bonding interactions consistently accentuate the effects of the substituent groups of the benzoquinones. This study provides direct support for the tuning and control of quinone cofactors in biological solar energy transduction through interactions with the surrounding protein matrix. Introduction Quinones are naturally occurring isoprenoids that are widely exploited by photosynthetic reaction centers (RC) (e.g., the bacterial RC,1-4 photosystem II,5-8 and photosystem I9-11). It is thought that protein interactions modify the properties of quinones such that similar quinone species can perform diverse functions in RCs.10-13 For example, both type I and type II (oxygenic and non-oxygenic) RCs contain quinone cofactors that serve very different functions, as the redox potential of similar quinones can operate at up to 800 mV lower reduction potential when present in type I RCs. Quinones also display an enormous versatility of function in energy-transducing respiratory protein complexes (e.g., cytochrome bc114-16 and fumarate reductase17). However, the factors that determine quinone function in energy transduction remain unclear. It is thought that the structure and substituent groups of the quinone, the location of the quinone cofactor, the geometry of its binding site, the redox potential, and “smart” matrix effects from the surrounding protein environment greatly influence the functional properties of quinones. Understanding the influence of structural and environmental effects on the electronic properties of quinones will help elucidate the tuning and control of quinone function in biological solar energy transduction. Previous studies have reported a diverse range of electrochemical redox potentials for quinones in photosynthetic protein complexes. In the type I reaction center, photosystem I (PSI), phylloquinones are used as intermediate electron acceptors in the electron transport chain.9-11,18 The range of redox potentials * To whom correspondence should be addressed. Phone: (518) 276 3271. Fax: (518) 276 4887. E-mail: [email protected].

for the phylloquinones of PSI is estimated to be between -800 and -700 mV.11,19,20 In the type II reaction centers, namely photosystem II (PSII) and the bacterial reaction center (BRC), plastoquinone, ubiquinone, and menaquinone function as terminal electron acceptors, respectively.1-8 In the BRC, the redox potential of the primary quinone acceptor, QA, is reported to be fairly high, between -45 and -75 mV,21,22 while the secondary quinone acceptor, QB, is reported to have a first redox potential of +40 mV and a second redox potential of -40 mV.22 For PSII, electrochemical measurements of the primary electron acceptor, QA, have yielded a range of redox potentials from -300 to +100 mV, depending on the sample conditions used in the study. In general, the redox potential of the primary quinone of PSII is estimated to be approximately -100 mV.18,23 Through the analysis of the primary sequence of the quinone binding pockets of photosynthetic protein complexes, Fisher and Rich have suggested several conserved amino acid residues that could potentially interact with the bound quinone molecules and influence their function.24 However, despite similarities of the quinone binding sites, there is no consensus on specific protein interactions that directly relate structure to function. The significance of protein interactions on quinone function is underscored in PSII, where the effects of the surrounding protein environment govern the vast differences in the function of the primary and secondary quinone electron acceptors. The primary and secondary quinone acceptors of PSII are identical plastoquinone molecules, with the primary quinone acting strictly as a single-electron acceptor and the secondary quinone acting as a two-electron/two-proton acceptor.25-28 This difference in the function of the primary and secondary plastoquinone acceptors of PSII is most likely due to interactions with the surrounding

10.1021/jp907379d CCC: $40.75  2009 American Chemical Society Published on Web 10/16/2009

15410

J. Phys. Chem. B, Vol. 113, No. 46, 2009

protein environment of the quinone binding pocket. This hypothesis is supported by a recent study by Golbeck and coworkers demonstrating the effects of asymmetric hydrogen bonding and π-stacking interactions on the half-wave reduction potential of a series of phylloquinone molecules.29 The significance of the influence of the surrounding protein environment on the functional tuning of quinone electron acceptors is also demonstrated by experiments in which the phylloquinone molecules of PSI were biosynthetically replaced by plastoquinone molecules.30-32 The analysis of these “nonnative” plastoquinones in PSI, which are estimated to have a -100 mV potential in the QA binding pocket of PSII, reveals a dramatic reduction of the redox potential to -650 mV due to the influence of the protein environment of the quinone binding pocket of PSI.33 The -650 mV redox potential of plastoquinones in PSI mirrors the -800 mV potential that is typically observed for the native phylloquinone acceptor of PSI, indicating that there are inherent interactions in the protein binding pocket that drastically lower the redox potential of the bound quinone acceptor.18 In addition to examining the functional tuning of quinones by sequence analysis and quinone substitution in reaction centers, there have also been investigations of the correlation of absorption spectra with the reduction potentials of quinones,34 the redox behavior and interactions of quinones with surrounding media,35-37 and the effects of hydrogen bonds on model quinones.38-40 In order to investigate the changing response of the environment as the quinones cycle through various redox states, electrostatic calculations have been performed of the BRC to investigate the redox potentials41 and the influence of hydrogen bonds on the redox potentials.42 Quantum mechanical and molecular modeling (QM/MM) methods have been used to examine the influence of the protein environment on the spin density distribution of quinones43 and the effect of hydrogen bonding interactions.44 In addition, Warncke, Gunner, and coworkers have conducted studies to compare the binding free energies of a series of quinones45 and the stability of the quinone redox states46 by continuum electrostatics calculations. Several spectroscopic techniques have been used to characterize proteinquinone interactions, including Fourier transform infrared spectroscopy (FTIR),47,48 continuous wave (CW)22 and pulsed electron paramagnetic resonance (EPR) spectroscopy studies.49-51 EPR spectroscopy is a powerful structural tool that can provide a direct measure of the magnetic interactions, local structure, bonding, electron spin densities, and the spatial orientation of paramagnetic centers.52,53 In liquids, rapid rotational reorientation makes room temperature CW EPR spectroscopy highly sensitive to the isotropic component of the electron-nuclear hyperfine interaction. The anisotropic components of the electron-nuclear hyperfine tensor can be obtained by analysis of EPR spectra in the solid state or in a frozen solution. However, weak hyperfine couplings are often obscured in the solid state due to inhomogeneous line broadening. Pulsed EPR spectroscopy methods, including one-dimensional electron spin echo envelope modulation (1D ESEEM) and twodimensional (2D) hyperfine sublevel correlation spectroscopy (2D HYSCORE or 2D ESEEM), are immensely useful for identifying protein structure by recovering weak anisotropic hyperfine interactions.53-56 In comparison with 1D ESEEM spectroscopy, the use of 2D HYSCORE enhances resolution by separating overlapping nuclear frequencies in a second dimension and simplifies analysis by correlating the nuclear frequencies that arise from different electron spin manifolds. The direct determination of the strength and the nature of the

Weyers et al. electron-nuclear couplings of quinones by a combination of CW and pulsed 2D HYSCORE EPR spectroscopy methods provide snapshots of the interactions inherent to the quinone and the interactions with the surrounding solvent environment that influence the electronic structure and hence the function of quinones. In the present study, we conduct a systematic investigation of a series of singly substituted, 2,5-disubstituted, and tetrasubstituted 1,4-benzoquinone model compounds and plastoquinone (PQ). We utilize CW and pulsed 2D HYSCORE spectroscopy to measure the isotropic and anisotropic hyperfine couplings, respectively, of the benzoquinone molecules in aprotic and protic solvents. The analysis of the combination of isotropic and anisotropic hyperfine couplings provides a direct measure of the electron spin density distribution on the individual atoms of the reduced benzosemiquinone state. We isolate the effect of hydrogen bonds on the electron spin density distribution by comparison of the CW and pulsed EPR spectra benzosemiquinones in aprotic and protic solvents. In parallel, we demonstrate the use of differential pulsed voltammetry (DPV) measurements to provide an estimate of redox potentials of the benzoquinone model compounds. We correlate the electron spin densities of the benzosemiquinones obtained by CW EPR and 2D HYSCORE spectroscopy with the substituent and solvent effects and correlate these effects with the observed redox potentials of the quinone model compounds. This study provides valuable insight on the influence of molecular interactions on the tuning and function of quinones in vitro, and it contributes to the elucidation of the control and tuning of quinones in biological solar energy transduction. Materials and Methods Preparation of Benzosemiquinone Anion Radicals. 1,4Benzoquinone (BQ) (98%), 2,5-dimethyl-1,4-benzoquinone (DMBQ) (99%), 2,3,5,6-tetramethyl-1, 4-benzoquinone (TMBQ) (99%) and 2-phenyl-1,4-benzoquinone (pPBQ) (99%) were purchased from Acros Organics (Morris Plains, NJ). 2,3,5,6Tetrachloro-1,4-benzoquinone (TCBQ) (99%) was purchased from Sigma-Aldrich (St. Louis, MO). 2,5-Dichloro-1,4-benzoquinone (DCBQ) (98%) and 2,5-diphenyl-1,4-benzoquinone (DPBQ) (99%) were purchased from Pfaltz & Bauer (Waterbury, CT). All quinones were obtained commercially and were used without further purification. The semiquinone anion radicals were generated by dissolving the benzoquinones (0.5-30 mM) in dimethyl sulfoxide (DMSO) or isopropyl alcohol (IPA) purged with argon gas. To this mixture, a 10 M solution of sodium hydroxide (aqueous) solution was added dropwise (100-fold molar excess) in an inert argon atmosphere. For samples dissolved in deuterated DMSO, the semiquinone anion radicals were generated using sodium deuteroxide (NaOD). The formation of the semiquinone anion radical is usually accompanied by a distinct color change of the mixture. The DMSO and IPA solutions of the semiquinone anion radicals were loaded into 2 mm glass capillaries for room temperature EPR spectroscopy measurements and the EPR spectra were acquired immediately. For cryogenic pulsed X-band 2D HYSCORE spectroscopy measurements, the samples were loaded into 4 mm quartz tubes (Wilmad Glass). Once loaded into the quartz tubes, the samples were further purged with argon gas and shock frozen at 77 K. Unless otherwise noted, the 1H 2D HYSCORE spectra were recorded in deuterated solvent to eliminate the electron-nuclear couplings with matrix (solvent) protons and enhance the spectral resolution.

Spectroscopic Descriptions of Model Quinones Electrochemistry Measurements. Differential pulse voltammetry (DPV) measurements were performed on a timeresolved electrochemical quartz microbalance CH440A (CH Instruments, Austin, TX). The samples were analyzed using a three-electrode system consisting of a glassy carbon working electrode, a platinum wire counter electrode, and a silver wire reference electrode. The quinone model compounds were scanned in dry acetonitrile with 0.1 M supporting electrolyte, tetrabutylammonium hexafluorophosphate (TBAP6) (98%, Acros Organics, Morris Plains, NJ). The DPV measurements were performed in a nitrogen atmosphere at room temperature, with ferrocene (98%, Sigma-Aldrich, St. Louis, MO; purified by sublimation) as an internal standard. All of the measurements are referenced to the ferrocene/ferrocinium (Fc/Fc+) internal standard and corrected to the normal hydrogen electrode (NHE) (Fc ) 0.40 V vs aqueous SCE, SCE ) -0.24 V vs NHE).57,58 CW and Pulsed EPR Spectroscopy. The EPR spectra were recorded on a custom-built CW/pulsed X-band Bruker Elexsys 580 EPR spectrometer. The CW EPR measurements were conducted at room temperature with a dual-mode resonator ER 4116-DM (Bruker BioSpin, Billerica, MA) equipped with a continuous-flow helium E900 cryostat (Oxford Instruments, Oxfordshire, UK). The operating microwave frequency was 9.64 GHz. The EPR spectra were acquired with a modulation frequency of 100 kHz and a modulation amplitude of 0.050.1 G. Typically 4 scans were acquired for each spectrum. The pulsed EPR measurements were conducted with a dielectric flex-line probe ER 4118-MD5 (Bruker BioSpin, Billerica, MA) and a dynamic continuous-flow cryostat CF935 (Oxford Instruments, Oxfordshire, UK). The operating microwave frequency of the pulsed resonator was 9.72 GHz, and the 2D HYSCORE spectra were recorded at the field position of maximum echo intensity (3460 G). The pulsed EPR spectra of the quinone models using DMSO as a solvent were recorded at 40 K and the pulsed EPR spectra with IPA as a solvent were recorded at 115 K. The temperatures for recording the spectra were chosen based on the best available compromise between the spin relaxation rates of the benzosemiquinone radicals and the size of the Boltzmann polarization. For the 2D HYSCORE spectra, the pulsed echo amplitude was measured using the sequence π/2-τ-π/2-t1-π-t2-π/ 2-echo with a τ of 132 ns and a 12 ns detector gate (centered at the maximum of the echo signal); the delays are defined as the differences in the pulse starting points. The echo intensity was measured as a function of t1 and t2, where t1 and t2 were incremented in steps of 8 or 16 ns from their initial values of 24 and 40 ns, respectively. Equal amplitude pulses of 16 ns for π/2 and 32 ns for π were used to record a 256 × 256 matrix. The 16 ns time difference between the initial values of t1 and t2 and π/2 and π were set equal to obtain more symmetric spectra. The unwanted echoes and antiechoes were eliminated by applying a 16-step phase cycling procedure.59 Data Processing and Simulations. The differential pulse voltammograms were plotted in Matlab R2008a where the recorded redox potentials were calibrated by referencing the Fc/ Fc+ internal standard to NHE standard potential. The experimental CW EPR spectra acquired at room temperature were simulated using the GARLIC EPR spectral simulation subroutine of EasySpin 3.0.0 designed reproduce isotropic CW EPR spectra on Matlab R2008a.60 The GARLIC subroutine allows for the simulation of EPR spectra involving electron-nuclear hyperfine couplings to multiple nuclei. For the 2D HYSCORE data, the echo decay was eliminated by loworder polynomial baseline correction and tapered with a

J. Phys. Chem. B, Vol. 113, No. 46, 2009 15411

Figure 1. Differential pulsed voltammograms of benzoquinone model compounds with ferrocene added as an internal standard. (A) The voltammograms for (a) BQ, (b) PQ, (c) pPBQ, and (d) DPBQ. (B) The voltammograms for (a) DMBQ, (b) TMBQ, (c) DCBQ, and (d) TCBQ. The redox potential of the unsubstituted 1,4-benzoquinone is marked by a dotted line, allowing for a qualitative visual comparison of substituent effects on redox potential. Potentials have been corrected to reflect calibration vs NHE.

Hamming function. Before 2D Fourier transformation, the data was zero filled to a 1024 × 1024 matrix and the magnitude spectra were calculated using the Bruker X-EPR software (Bruker BioSpin, Billerica, MA). The spectra are presented as contour plots prepared in Matlab R2008a. Results Shown in Figure 1 are the differential pulsed voltammetry (DPV) measurements of the benzoquinone model compounds with ferrocene added as an internal standard. Figure 1A displays the pulsed voltammograms of BQ, PQ, pPBQ, and DPBQ and Figure 1B displays the pulsed voltammograms of DMBQ, TMBQ, DCBQ, and TCBQ. As can be seen, the DPV measurements yield a peak output (rather than the cyclic wave obtained by cyclic voltammetry measurements). In each case, the peak marks the first redox potential for the quinone/semiquinone redox couple. In addition, a ferrocene peak is present, used as an internal calibration standard.61 The redox potential of unsubstituted 1,4-benzoquinone is marked by a dotted line in both set of voltammograms shown in Figure 1A,B, allowing for a qualitative visual comparison of substituent effects on redox potential. As can be seen, the presence of electrondonating substituent groups on the benzosemiquinone decrease the redox potential (i.e., a more negative value) while electronwithdrawing groups increase the redox potential (i.e., a more

15412

J. Phys. Chem. B, Vol. 113, No. 46, 2009

Weyers et al.

TABLE 1: First Redox Potential of Benzoquinone Model Compounds Calibrated vs NHE in Dry Acetonitrile with TBAP6 Supporting Electrolyte quinone (abbrev)

redox potentiala (V)

1,4-benzoquinone (BQ) plastoquinone (PQ) 2-phenyl-1,4-benzoquinone (pPBQ) 2,5-diphenyl-1,4-benzoquinone (DPBQ) 2,5-dimethyl-1,4-benzoquinone (DMBQ) tetramethyl-1,4-benzoquinone (TMBQ) 2,5-dichloro-1,4-benzoquinone (DCBQ) tetrachloro-1,4-benzoquinone (TCBQ)

-0.043 -0.351 -0.034 -0.025 -0.210 -0.385 0.227 0.476

a

Corrected vs NHE.

Figure 3. CW X-band EPR spectra of the benzosemiquinone anion radicals acquired with IPA as the solvent. The EPR spectra of (A) (a) BQ, (b) PQ, and (c) DPBQ. (B) The EPR spectra of (a) DMBQ, (b) TMBQ, (c) DCBQ, and (d) TCBQ. The experimental spectra are represented as solid lines, and the corresponding spectral simulation is shown as a dashed line.

Figure 2. CW X-band EPR spectra of the benzosemiquinone anion radicals acquired with DMSO as the solvent. (A) The EPR spectra of (a) BQ, (b) PQ, (c) pPBQ, and (d) DPBQ. (B) The EPR spectra of (a) DMBQ, (b) TMBQ, (c) DCBQ, and (d) TCBQ. The experimental spectra are represented as solid lines, and the corresponding spectral simulation is shown as a dashed line.

positive value) relative to the unsubstituted 1,4-benzosemiquinone. The quantitative values of the first redox potential, which is the conversion of the neutral benzoquinone to the oneelectron-reduced benzosemiquinone anion radical, are obtained from the peak position of the redox couple in each voltammogram and are presented (calibrated vs NHE) in Table 1. Figures 2A,B and 3A,B show the CW X-band experimental (solid line) and simulated (dashed line) EPR spectra of the model benzosemiquinone anion radicals in an aprotic solvent, DMSO (Figure 2), and in a protic solvent, IPA (Figure 3), respectively. The CW EPR spectra of liquids at room temperature provide a sensitive method for assessing the directionally independent “isotropic” component of the electron-nuclear hyperfine couplings (hfcs), aiso, that is a result of the contact interaction. The observation of the isotropic hfcs is facilitated by the rapid averaging of the directionally dependent “anisotropic” components of the hfcs resulting from the dipolar interaction and by the narrow EPR spectral line widths of the benzosemiquinones in solution. As seen in Figures 2 and 3, the benzosemiquinone EPR spectra are characterized by a distinctive pattern of

hyperfine splittings. The number, the positions, and the relative intensities of the peaks in the EPR spectra are governed by the strength of interactions (aiso), the number (n), and the spin state (I) of the nuclei that are magnetically coupled to the electron spin. The hyperfine couplings observed in this study arise from the nuclear spin (I ) 1/2) of protons. The hyperfine splitting patterns displayed by the benzosemiquinone models are numerically simulated to obtain an excellent agreement between the experimental and simulated EPR spectra. The simulated isotropic hyperfine coupling constants of the benzosemiquinone anion radicals are presented in Table 2. For nonequivalent protons, the isotropic hfcs are assigned based on simulation constraints or to qualitatively match values in literature.62,63 For pPBQ (in DMSO), it is reasonable to postulate that each proton on the quinone ring would have a distinct magnetic interaction with the unpaired electron density of the quinone and thus yield distinct hfcs. However, these hfcs were not resolved in the present study, although the hyperfine couplings of the phenyl protons in DPBQ and pPBQ were resolved. Figure 4 displays the 2D HYSCORE spectra of the 1,4benzosemiquinone anion radicals in deuterated IPA. In the 2D HYSCORE spectra, the electron-nuclear hyperfine interactions appear in different quadrants of the spectrum depending on the strength of the electron-nuclear hyperfine couplings. For the benzosemiquinone anion radicals investigated in this study, there is no appreciable intensity of cross peaks observed in the (+,-) quadrant, making the (+,+) quadrant the only nondegenerate quadrant of interest. It is known that the cross peaks arising from weakly interacting nuclei will appear in the (+,+) quadrant of the 2D HYSCORE spectra and the cross peaks are centered about the diagonal located at the Zeeman frequency of the interacting nuclei.59 As seen in Figure 4A, the 2D HYSCORE

Spectroscopic Descriptions of Model Quinones

J. Phys. Chem. B, Vol. 113, No. 46, 2009 15413

TABLE 2: Isotropic Hyperfine Coupling Constants and Electron Spin Densities Obtained from the Spectral Simulations of the CW EPR Spectra of Benzosemiquinone Models in Aprotic and Protic Solvents, DMSO and IPA, Respectively CW EPR parameters (IPA) aiso (MHz)

quinone (abbrev) 1,4-benzoquinone (BQ) plastoquinone (PQ) 2-phenyl-1,4-benzoquinone (pPBQ) 2,5-diphenyl-1,4-benzoquinone (DPBQ) 2,5-dimethyl-1,4-benzoquinone (DMBQ) tetramethyl-1,4-benzoquinone (TMBQ) 2,5-dichloro-1,4-benzoquinone (DCBQ) tetrachloro-1,4-benzoquinone (TCBQ)

6.62 4.86 6.28 6.20 5.3 5.70 na

spin density (%)

(a), 5.40 (b), 6.70 (c), 5.76 (d) (R2,R4) 0.3, 0.3, 0.75a (R1,R3), 5.20 (R2,R4) (R2,R4)

8.83 8.10 (a), 9.00 (b), 7.44 (c), 7.68 (d) 8.37 (R2,R4) 10.33 (R1,R3), 6.93 (R2,R4) 8.83 7.60 (R2,R4) na

CW EPR parameters (DMSO)

a

quinone (abbrev)

aiso (MHz)

spin density (%)

1,4-benzoquinone (BQ) plastoquinone (PQ) 2-phenyl-1,4-benzoquinone (pPBQ) 2,5-diphenyl-1,4-benzoquinone (DPBQ) 2,5-dimethyl-1,4-benzoquinone (DMBQ) tetramethyl-1,4-benzoquinone (TMBQ) 2,5-dichloro-1,4-benzoquinone (DCBQ) tetrachloro-1,4-benzoquinone (TCBQ)

6.71 5.29 (a), 5.97 (b), 4.95 (c), 4.86 (d) 6.8 (R4) 6.1 (R2,R3) 0.34, 0.66, 1.18a 6.67 (R2,R4) 0.30, 0.41, 1.12a 6.04 (R1,R3), 5.53 (R2,R4) 5.38 6.16 (R2,R4) na

8.95 8.82 (a), 9.95 (b), 5.50 (c), 6.48 (d) 9.07 (R4), 8.13 (R2,R3) 8.89 (R2, R4) 10.07 (R1,R3), 7.37 (R2,R4) 8.97 8.21 (R2,R4) na

Phenyl ring protons.

Figure 4. 2D HYSCORE spectra of the 1,4-benzosemiquinone anion radical in deuterated IPA: (A) the full 2D HYSCORE spectrum and (B) the enlargement of cross peaks in the (+,+) quadrant that are centered at the proton Larmor frequency of 15 MHz, (C) proton cross peaks plotted in the frequency-squared coordinate system, and (D) the simulation of cross peaks using parameters obtained from best spectral fit of the experimental 2D HYSCORE spectrum.

spectrum of the BQ anion radical in deuterated IPA (IPA-d8) in the (+,+) quadrant is dominated by the solvent matrix deuterium cross peaks lying on the deuteron Zeeman frequency diagonal (2.3 MHz) and its multiples. The proton cross peaks that are of interest in this study (shown in Figure 4B) are centered in the spectral region at the proton Larmor frequency of 14.7 MHz. The 2D HYSCORE proton cross peaks were analyzed to obtain a measure of the anisotropic hyperfine couplings of the benzosemiquinone anion radicals using the following procedure based on the method developed by Dikanov et al.64 In general, a hyperfine tensor is characterized by its three principal

components and the corresponding Euler angles that define the orientation of the hyperfine tensor in the molecular frame. Under the given experimental conditions for the benzosemiquinone radicals in a frozen solution, all of the molecular orientations of the tensor are excited with nearly equal probability (there is no significant orientation selection at X-band EPR frequency) and therefore only the principal values of the hyperfine tensor are observed from the experimental 2D HYSCORE spectra. Assigning aiso, T, and δ as the isotropic, dipolar, and rhombic components of the hyperfine coupling, respectively, the principal values of a hyperfine tensor can be represented as aiso - T(1 - δ), aiso - T(1 + δ), aiso + 2T with 0 e δ e 1. In the

15414

J. Phys. Chem. B, Vol. 113, No. 46, 2009

Weyers et al.

present analysis, we assume axial symmetry for the hyperfine tensors, i.e., δ ) 0. In this case, the hyperfine interactions are determined by two components, aperp, where the direction of magnetic field is perpendicular to the symmetry axis, and apara, where the direction of magnetic field is parallel to the symmetry axis. The correlation patterns represent a segment of an ellipse and are determined by following equations:65 2 VR(β) ) (QR(β)Vβ(R) + GR(β))1/2

(1)

where

QR(β) )

2aiso + T - 4V1 2aiso + T ( 4V1

and

GR(β) ) (2VI

2 4V2I - aiso + 2T2 - aisoT 2aiso + T ( 4VI

(2)

As suggested by Dikanov and co-workers, the analysis of the 2D HYSCORE data can be simplified if the spectra are plotted in squared-frequency coordinates.65 Here, the ridges of the cross peaks are straight line segments with the slope and intercept represent QR(β) and GR(β), respectively. The simple transformation of eq 2 provides direct expressions for the hyperfine components:

T)(

and



{

4VI2QR(β) 16 GR(β) + 9(1 - QR(β) 1 - QR(β)

}

(3)

aiso ) (2VI

1 + QR(β) T 1 - QR(β) 2

(4)

To obtain reliable values for QR(β) and GR(β), only points from the middle of the highest level, narrow-featured contours were considered. Taking points from the middle of the ridges (or cross peaks) is justified by the fact that most of the broadening effects that contribute to the width of the experimental ridges are isotropic in the 2D frequency space and this will not shift the position of the ridge. It is also known that the benzosemiquinone ring protons exhibit small rhombicity.66 However, this does not significantly affect the quality of our analysis, since for small values of δ, the ridges in the squared-frequency space become filled triangles. The slope and intercept of the median of such a triangle in the first order of approximation are the same as the position and direction of the straight line ridges produced when δ ) 0. As shown in Figure 4C, plotting the 2D HYSCORE spectra of the 1,4-benzosemiquinone anion radical in the squaredfrequency coordinate system straightens the ridges; points taken from the middle of this ridge can be measured and fit to a straight line with good agreement. From the slope and intercept of this line the hfc parameters aiso and T were calculated using eqs 4. These parameters were then used to simulate the 2D HYSCORE data, as seen in Figure 4D. Although there are minor deviations in the intensity patterns along the ridges, the positions of the ridges are well reproduced by the simulation shown in Figure 4D. The main reason for minor deviations of the intensity patterns in the spectral simulations is the nonideality (nonzero time duration) of the applied experimental microwave pulses, which are assumed to be ideal in the simulations. The strength of the microwave field was V1 ≈ 15 MHz, which is far too small to satisfy the assumed simulation conditions, i.e., V1 . |VI, aiso, T|. While the nonideality of the pulses affects intensity distribution within the cross peaks, it does not change the cross peak position. Thus, it does not affect the validity of our analysis. Representative experimental 2D HYSCORE spectra of benzosemiquinone anion radicals in DMSO and IPA are shown in

Figure 5. Representative experimental 2D HYSCORE spectra of additional benzosemiquinone anion radicals using DMSO and IPA as the solvents. Shown are the experimental 2D HYSCORE spectra of 2,5-dimethyl-1,4-benzosemiquinone in (A) DMSO-d6 and (B) IPA-d6 and the tetrachloro1,4-benzosemiquinone in (C) DMSO and (D) IPA.

Spectroscopic Descriptions of Model Quinones

J. Phys. Chem. B, Vol. 113, No. 46, 2009 15415

TABLE 3: Isotropic Hyperfine Coupling Constants, Electron Spin Densities, and Anisotropic Hyperfine Coupling Constants Obtained from the Spectral Simulations of the 2D HYSCORE Spectra of Benzosemiquinone Models in Aprotic and Protic Solvents, DMSO, and IPA, Respectively HYSCORE parameters (IPA) quinone (abbrev)

aiso (MHz)

spin density (%)

aperp (MHz)

apara (MHz)

T (MHz)

1,4-benzoquinone (BQ) 2-phenyl-1,4-benzoquinone (pPBQ) 2,5-diphenyl-1,4-benzoquinone (DPBQ) 2,5-dimethyl-1,4-benzoquinonea (DMBQ) tetramethyl-1,4-benzoquinoneb (TMBQ) 2,5-dichloro-1,4-benzoquinone (DCBQ) tetrachloro-1,4-benzoquinone (TCBQ)

6.92 6.50 6.34 5.98 (a), 5.36 (b) 5.6 5.81 0.44

9.23 8.67 8.45 9.96 (a), 7.14 (b) 9.30 7.75 -

10.01 9.50 9.19 4.36 (a), 8.12 (b) 4.4 8.61 -2.50

0.75 0.51 0.64 9.21 (a), -0.18 (b) 7.9 0.21 6.34

-3.08 -3.00 -2.85 1.62 (a), -2.77 (b) 1.17 -2.80 2.95

HYSCORE Parameters (DMSO) quinone (abbrev)

aiso (MHz)

spin density (%)

aperp (MHz)

apara (MHz)

T (MHz)

1,4-benzoquinone (BQ) 2-phenyl-1,4-benzoquinone (pPBQ) 2,5-diphenyl-1,4-benzoquinone (DPBQ) 2,5-dimethyl-1,4-benzoquinonea (DMBQ) tetramethyl-1,4-benzoquinone (TMBQ) 2,5-dichloro-1,4-benzoquinone (DCBQ) tetrachloro-1,4-benzoquinone (TCBQ)

6.81 6.57 6.54 5.70 (a), 5.43 (b) 5.41 6.18 na

9.07 8.76 8.71 9.49 (a), 7.23 (b) 9.01 8.24 -

9.80 9.35 9.25 4.11 (a), 8.06 (b) 3.96 8.81 -

0.82 1.00 1.11 8.86 (a), -0.16 (b) 8.29 0.91 -

-2.99 -2.78 -2.71 1.58 (a), -2.63 (b) 1.44 -2.63 -

a

Hfc parameters are designated as methyl group protons (a) and ring protons (b). b Parameters from W-band ENDOR(38).

Figure 5. Shown in Figure 5A,B are the experimental 2D HYSCORE spectra of the 2,5-dimethyl-1,4-benzosemiquinone in DMSO and IPA, respectively. Also shown are the experimental 2D HYSCORE spectra of the tetrachloro-1,4-benzosemiquinone in DMSO (Figure 5C) and in IPA (Figure 5D). Similar to the 1,4-benzosemiquinone spectrum in Figure 4A,B, we observe a single proton ridge in each cross peak for the benzosemiquinone radical anions except for those arising from plastoquinone (data not shown) and DMBQ (Figures 5A,B). The hyperfine coupling constants obtained from the simulations of the 2D HYSCORE spectra are shown in Table 3. The isotropic hfcs obtained from the 2D HYSCORE spectra are in good agreement with the hfc constants obtained from the room temperature CW EPR spectra simulations (Table 2). As previously mentioned, pPBQ, PQ, and DMBQ have more than one group of equivalent protons, each having distinct values of hfcs. In DMBQ these couplings are separated into two distinct ridges; however, in PQ and pPBQ the hfcs were not resolved (data not shown). In PQ, the lack of resolution is most likely explained by (a) the greater number of proton groups with closely spaced cross peaks and (b) the use of protonated solvent, which results in a significant increase in the intrinsic line widths. In pPBQ, in principle each of the ring protons should exhibit a different electron-nuclear interaction. However, the couplings are nearly equivalent and are not resolved in either the room temperature CW EPR spectra or in the 2D HYSCORE spectra. The acquisition of 2D HYSCORE spectra of TCBQ in aprotic (DMSO) and protic (IPA) solvents presents a unique opportunity to examine the interaction between solvent protons and the benzosemiquinone anion radicals as all of the proton hyperfine interactions observed in the 2D HYSCORE spectrum of TCBQ must arise from the interactions with the solvent. As seen in Figure 5C,D, there is clear interaction between the TCBQ semiquinone radical and the protonated solvent. The observed ridges for TCBQ due to solvent hfcs are significantly broader than the ridges that are observed from the ring protons on the other model benzoquinones in deuterated solvents, due to the inhomogeneous broadening in protonated solvents. However, analysis of the two sets of ridges observed in IPA indicates that they are well extended and are shifted substantially from the antidiagonal VR + Vβ ) 2VI, which indicates a significant

anisotropic interaction, typical for hydrogen bonds. The calculated values for the hydrogen bond hfcs are in good agreement with other hydrogen bond hfcs previously reported (Table 3).40 The hydrogen bond hfcs measured in water reported by Sinnecker et al. show that there is minimal change in hydrogen bond hfcs from water to IPA.39 In DMSO, while the total span of the ridges is similar to those measured in IPA, significant differences are observed in the interaction. First, the peak position is centered on the antidiagonal, with a smaller shift, indicating either much weaker interaction anisotropy or the partial averaging of several different anisotropies over a broader distribution of nuclear interactions. This broad distribution could arise either from having a large set of proton interactions with each benzosemiquinone or from a large set of different interactions among all of the benzosemiquinones in the sample. Second, when comparing the intensities of the ridges, it can be seen that the intensities of the cross peaks arising from interactions with DMSO decay rapidly for separations of more than 2.1 MHz, while the IPA cross peak intensities are much more widely spaced and broadly distributed. This indicates that the DMSO interactions are weaker than those in IPA. From these two differences and the similarity of our IPA spectra with other reported hydrogen-bonding parameters,40 we conclude that the IPA 2D HYSCORE spectra are a result of hydrogen-bonding interactions with the solvent protons while the DMSO spectra likely result from very weak (nonspecific) interactions of the solvent matrix protons. Discussion The redox potential of the quinone moiety is a measure of the electron affinity that directly reflects the ease of reduction of the quinone system. Thus, it is expected that the addition of electron density to the molecule through an electron-donating substituent groups would decrease the electron affinity, creating a negative shift in redox potential, while the presence of electron-withdrawing substituent groups would increase electron affinity and induce a positive shift in the redox potential. As can be seen in Table 1, the redox potentials for the series of benzoquinone model compounds range over ∼1.0 V, from -0.385 to 0.476 V. The examination of the trend of redox

15416

J. Phys. Chem. B, Vol. 113, No. 46, 2009

Weyers et al.

potentials outlined in Table 1 confirms that the presence of electron-donating groups shift the redox potential to progressively negative values while the presence of electron-withdrawing groups causes positive shifts of the redox potential. Further, examination of the difference between the singly and disubstituted (and the disubstituted and tetrasubstituted) benzoquinone model compounds reveals that doubling the number of groups roughly doubles the shift in redox potential relative to the unsubstituted 1,4-benzoquinone, and this trend is observed in all substituent groups investigated in this study. For conjugated π-systems such as benzoquinones, the conjugated system as a whole is affected by the presence of an unpaired electron spin, leading to a distribution of unpaired electron spin density across the molecule. In conjugated systems, the observed hfcs associated with a particular nuclear spin are proportional to the spin density on the ring carbon atom adjacent to that nucleus, typically a proton, in the aromatic system. The spectral simulations of the experimental CW EPR spectra shown in Figures 2 and 3 provide the isotropic hfcs associated with the electron-nuclear couplings with ring protons and substituent group protons, providing a picture of the spin density on the adjoining quinone ring carbons.67 Listed in Table 2 are the isotropic hfcs obtained from the spectral simulation of the room-temperature CW EPR spectrum of each benzosemiquinone anion radical. The related spin densities can be calculated for different types of ring substituents using the McConnell equation. The McConnell equation for the calculation of spin density adjacent to R-protons68 is as follows:

aN ) QFn

(5)

where aN is the hfc for N nucleus, FN is the spin density at the R-carbon adjacent to nucleus N, and Q is a constant. For 1,4benzoquinones, Q is known with a value of -75 MHz.69 For methyl protons, a simplified version of the McLachlan relationship can be used:62

aN ) (B1 /2)FN

(6)

where B1 is a known constant, normally taken to be 120 MHz for 1,4-benzoquinones.62,69 For PQ, the original McLachlan relationship is used to calculate the spin densities on carbon atoms adjacent to β-methylene protons, taking into account the angle the β-methylene protons make with the ring:70

aN ) (B0 + B1 cos2 θ)FN

(7)

where B0 is a constant normally approximated to zero.69 Here care must be extended to correctly define the angle made between the β-methylene protons (and thus the plastoquinone tail) and the quinone ring. Previous studies conducted by Zheng and Dismukes defined the angle between the β-methylene protons and the ring to be 30° for PQ in solution (IPA) and bound in the QA site of the PSII complex.62 By comparing the spin densities calculated from the CW EPR simulations for each model benzoquinone with the spin densities of the unsubstituted 1,4-benzoquinone, an understanding of the influence of each substituent on the overall electron density distribution of the quinone molecule can be developed. This provides a glimpse of the overall electronic properties of the benzoquinone. In the present study, the ring carbon atoms and

Figure 6. (A) Chemical structure and numbering system of model benzoquinones. R1 ) R2 ) R3 ) R4 ) H for BQ; R1 ) phenyl, R2 ) R3 ) R4 ) H for pPBQ; R1 ) R3 ) phenyl, R2 ) R4 ) H for DPBQ; R1 ) R3 ) Me, R2 ) R4 ) H for DMBQ; R1 ) R2 ) R3 ) R4 ) Me for TMBQ; R1 ) R3 ) Cl, R2 ) R4 ) H for DCBQ; R1 ) R2 ) R3 ) R4 ) Cl for TCBQ. (B) The chemical structure and numbering system for PQ.

substituent group positions are identified using the numbering system shown in Figure 6. In the phenyl-substituted semiquinone anions, the singly substituted compound shows an increase in spin density on carbon 5 adjacent to the phenyl group and a decrease in spin density on the opposite ring carbons 2 and 3. This is possibly due to the dual nature of the phenyl group, electron-withdrawing by resonance and electron-donating through inductive effects, reducing the overall electron density and the spin density across the molecule while the inductive effects result in a slight increase of the electron spin density on carbon 5, the adjacent ring carbon. The doubly substituted quinone shows a decrease in the spin density on both carbons (ring carbons 2 and 5). In the diphenyl-substituted quinone, the observed spin density on carbons 2 and 5 is close to the spin density distribution observed in the unsubstituted 1,4-benzosemiquinone. The overall effect of the methyl group substituents results in the increase of unpaired electron spin density on the benzosemiquinone (Table 2). This is consistent with the electron-donating nature of methyl group substituents. For the dimethyl-substituted quinone, the electron spin density on carbons 3 and 6 that are adjacent to the methyl group substituents displays an increase, while the electron spin density on carbons 2 and 5 decreases. This is possibly in response to perturbations in the overall ring density. For the tetramethyl-substituted semiquinone, the electron spin density is relatively unchanged compared to the unsubstituted 1,4-benzoquinone. For the chloro-substituted benzosemiquinone anion radicals, information on the dichloro-substituted quinone via the R-hydrogen hfcs can be determined. The spin density on carbons 2 and 5 of the benzosemiquinone exhibits a clear decrease in value (Table 2). This is consistent with an overall decrease of the electron density distribution across the ring, in agreement with the electron-withdrawing nature of the chloro substituent. Comparing the CW EPR spectra in IPA with those in DMSO reveals the same overall substituent trends for electron-donating and electron-withdrawing groups. The effect of a changing solvent environment on the model quinone series shows a tendency of the protic solvent (IPA) to exaggerate the influence of a substituent group. This can be seen in the case of DMBQ,

Spectroscopic Descriptions of Model Quinones where unpaired electron spin density increases on the carbon atoms (carbon atoms 3 and 6) adjacent to the methyl substituents by ∼13% in DMSO and 17% in IPA, while spin density decreases by 18% on the alternating carbon atoms (carbon atoms 2 and 5) in DMSO and 22% in IPA. The influence of the protic solvent is discussed further in analysis of the 2D HYSCORE data. The 2D HYSCORE experiment creates a two-dimensional observation of cross peaks correlated to the nuclear Larmor frequencies from the surrounding nuclear spins that are magnetically coupled to the electron spin observed in the experiment. The appearance and intensity of the cross peaks arising from the electron-nuclear interactions depend on the type of hyperfine interactions that are present. In the limit of weak hyperfine interactions, |T + 2aiso| , 4VI, where T is the perpendicular component of the axial anisotropic hyperfine tensor, aiso is the isotropic hyperfine coupling, and VI is the nuclear Larmor frequency, the cross peaks will appear in the (+,+) quadrant. In the limit of strong hyperfine interactions, |T + 2aiso| . 4VI the cross peaks appear in the (+,-) quadrant. When the hyperfine couplings are on the order of the nuclear Larmor frequency, the cross peaks will appear in both quadrants. For the benzosemiquinone anion radicals investigated in this study, the hyperfine couplings are within the weak coupling limit and appear in the (+,+) quadrant. There is no appreciable intensity that is observed in the (+,-) quadrant. Representative 2D HYSCORE spectra are shown in Figures 4, 5, and 6, and the hfcs determined from spectral analysis (as discussed in the Results section) are shown in Table 3. The hyperfine characteristics aiso and T for each nuclear hyperfine interaction have been determined from the 2D HYSCORE spectra acquired in this study. On the basis of the parameters determined by the spectral simulations, we calculated the spin density at the corresponding R-carbons and compared the results with the observed hfcs and electron spin densities obtained from the room temperature EPR spectroscopy studies. For the 2D HYSCORE measurements of the benzosemiquinone anion radicals in DMSO and IPA, the spin densities are tantamount to those measured by CW EPR spectroscopy, with the same trends and patterns of response to the ring substituent effects that were observed in the previous description. The comparison of the two sets of calculated spin densities for the 2D HYSCORE spectra reveals that the presence of the hydrogen bonds accentuates observed electronic variations; substituents that lower spin density are further decreased and those that increase spin density are further augmented. In IPA, it is expected that in the first solvent shell there would be four hydrogen bonds between the solvent and benzosemiquinone radical anion, two for each carbonyl moiety.71 The presence of hydrogen bonds should be the defining difference between the spectra acquired with DMSO and IPA as solvents. For example, in DMBQ the spin density on the ring carbons adjacent to the methyl groups is increased spin density (relative to the unsubstituted 1,4-benzoquinone) to 9.96% in IPA and 9.49% in DMSO (compared with BQ spin densities of 9.23% and 9.07%, respectively). The change in spin density is thus increased in IPA by 8% and 5% in DMSO. Similar calculations for DPBQ show a 9% decrease in IPA and 4% in DMSO, and for DCBQ a 16% decrease in IPA with a 9% decrease in DMSO. From these comparisons, it can be seen that the effect of a substituent group on the spin density of a model benzosemiquinone, whether it be to increase or decrease spin density, is accentuated in IPA. The effect of the solvating environment on the electronic properties of the 1,4-benzoquinone model compounds that are

J. Phys. Chem. B, Vol. 113, No. 46, 2009 15417 examined in this study, sometimes as much as doubling the influence of a particular substituent group, shows that the presence of hydrogen bonding is an important consideration in the tuning of the redox potential of the quinone cofactor. The analysis of the relative redox potentials and the electron spin density values obtained from CW EPR and 2D HYSCORE spectra of the semiquinone radical anions demonstrate the influence of both the substituent groups and the interactions with the surrounding environment. In order to understand the function of a specific quinone on the molecular level, a detailed understanding of the many factors which influence the electronic structure of the quinone is essential. This makes the real nature of the functional tuning of the quinone very complex. Thus, the development of an understanding of the effects of the chemical structure and hydrogen bonding on the electronic structure of quinone model systems is of great importance. In this study, we systematically measure the dependence of observables, such as the hyperfine interaction and redox potential, which are sensitive to the electronic structure of the semiquinone. This provides insight of the effects of chemical structure and solute-solvent interactions on the properties of model quinones, which contributes to an understanding of the functional tuning of quinones in photosynthetic reaction centers. Future studies will include a direct determination of the strength and nature of amino acid interactions in the primary and secondary quinone binding pockets of photosystem II. Acknowledgment. This research is supported by the Solar Energy Utilization Program, Office of Basic Energy Sciences, U.S. Department of Energy (DE-FG02-0ER06-15). The authors thank Professor Dinolfo for his valuable expertise and suggestions on the DPV measurements. References and Notes (1) Lancaster, C. R. D.; Michel, H. J. Mol. Biol. 1999, 286, 883. (2) Allen, J. P.; Feher, G.; Yeates, T. O.; Rees, D. C.; Deisenhofer, J.; Michel, H.; Huber, R. Proc. Natl. Acad. Sci. U.S.A. 1986, 83, 8589. (3) Feher, G.; Allen, J. P.; Okamura, M. Y.; Rees, D. C. Nature 1989, 339, 111. (4) Deisenhofer, J.; Epp, O.; Miki, K.; Huber, R.; Michel, H. Nature 1985, 318, 618. (5) Zouni, A.; Witt, H. T.; Kern, J.; Fromme, P.; Krauss, N.; Saenger, W.; Orth, P. Nature 2001, 409, 739. (6) Kamiya, N.; Shen, J. R. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 98. (7) Ferreira, K. N.; Iverson, T. M.; Maghlaoui, K.; Barber, J.; Iwata, S. Science 2004, 303, 1831. (8) Loll, B.; Kern, J.; Saenger, W.; Zouni, A.; Biesiadka, J. Nature 2005, 438, 1040. (9) Krauss, N.; Schubert, W. D.; Klukas, O.; Fromme, P.; Witt, H. T.; Saenger, W. Nat. Struct. Biol. 1996, 3, 965. (10) Jordan, P.; Fromme, P.; Witt, H. T.; Klukas, O.; Saenger, W.; Krauss, N. Nature 2001, 411, 909. (11) Srinivasan, N.; Golbeck, J. H. Biochim. Biophys. Acta-Bioenerg. 2009, 1787, 1057. (12) Rawls, R. Chem. Eng. News 2001, 79, 9. (13) Teutloff, C.; Bittl, R.; Stein, M.; Jordan, P.; Fromme, P.; Krauβ, N.; Lubitz, W. In 12th International Congress on Photosynthesis, Brisbane, Australia, 2001. (14) Iwata, S.; Lee, J. W.; Okada, K.; Lee, J. K.; Iwata, M.; Rasmussen, B.; Link, T. A.; Ramaswamy, S.; Jap, B. K. Science 1998, 281, 64. (15) Zhang, Z. L.; Huang, L. S.; Shulmeister, V. M.; Chi, Y. I.; Kim, K. K.; Hung, L. W.; Crofts, A. R.; Berry, E. A.; Kim, S. H. Nature 1998, 392, 677. (16) Xia, D.; Yu, C. A.; Kim, H.; Xian, J. Z.; Kachurin, A. M.; Zhang, L.; Yu, L.; Deisenhofer, J. Science 1997, 277, 60. (17) Iverson, T. M.; Luna-Chavez, C.; Cecchini, G.; Rees, D. C. Science 1999, 284, 1961. (18) Golbeck, J. H. Annu. ReV. Biophys. Biomol. Struct. 2003, 32, 237. (19) Brettel, K.; Leibl, W. Biochim. Biophys. Acta-Bioenerg. 2001, 1507, 100.

15418

J. Phys. Chem. B, Vol. 113, No. 46, 2009

(20) Ptushenko, V. V.; Cherepanov, D. A.; Krishtalik, L. I.; Semenov, A. Y. Photosynth. Res. 2008, 97, 55. (21) Dutton, P. L.; Leigh, J. S.; Wraight, C. A. FEBS Lett. 1973, 36, 169. (22) Rutherford, A. W.; Evans, M. C. W. FEBS Lett. 1980, 110, 257. (23) Krieger, A.; Rutherford, A. W.; Johnson, G. N. Biochim. Biophys. Acta-Bioenerg. 1995, 1229, 193. (24) Fisher, N.; Rich, P. R. J. Mol. Biol. 2000, 296, 1153. (25) Crofts, A. R.; Wraight, C. A. Biochim. Biophys. Acta 1983, 726, 149. (26) Velthuys, B. R.; Amesz, J. Biochim. Biophys. Acta 1974, 333, 85. (27) Kacprzak, S.; Kaupp, M. J. Phys. Chem. B 2004, 108, 2464. (28) Kern, J.; Renger, G. Photosynth. Res. 2007, 94, 183. (29) Feldman, K. S.; Hester, D. K.; Golbeck, J. H. Bioorg. Med. Chem. Lett. 2007, 17, 4891. (30) Johnson, T. W.; Shen, G. Z.; Zybailov, B.; Kolling, D.; Reategui, R.; Beauparlant, S.; Vassiliev, I. R.; Bryant, D. A.; Jones, A. D.; Golbeck, J. H.; Chitnis, P. R. J. Biol. Chem. 2000, 275, 8523. (31) Semenov, A. Y.; Vassiliev, I. R.; van der Est, A.; Mamedov, M. D.; Zybailov, B.; Shen, G. Z.; Stehlik, D.; Diner, B. A.; Chitnis, P. R.; Golbeck, J. H. J. Biol. Chem. 2000, 275, 23429. (32) Zybailov, B.; van der Est, A.; Zech, S. G.; Teutloff, C.; Johnson, T. W.; Shen, G. Z.; Bittl, R.; Stehlik, D.; Chitnis, P. R.; Golbeck, J. H. J. Biol. Chem. 2000, 275, 8531. (33) Shinkarev, V. P.; Zybailov, B.; Vassiliev, I. R.; Golbeck, J. H. Biophys. J. 2002, 83, 2885. (34) Peover, M. E. Trans. Faraday Soc. 1962, 58, 1656. (35) Wawzonek, S.; Berkey, R.; Blaha, E. W.; Runner, M. E. J. Electrochem. Soc. 1956, 103, 456. (36) Quan, M.; Sanchez, D.; Wasylkiw, M. F.; Smith, D. K. J. Am. Chem. Soc. 2007, 129, 12847. (37) Jaworski, J. S.; Lesniewska, E.; Kalinowski, M. K. J. Electroanal. Chem. 1979, 105, 329. (38) Rohrer, M.; MacMillan, F.; Prisner, T. F.; Gardiner, A. T.; Mobius, K.; Lubitz, W. J. Phys. Chem. B 1998, 102, 4648. (39) Sinnecker, S.; Reijerse, E.; Neese, F.; Lubitz, W. J. Am. Chem. Soc. 2004, 126, 3280. (40) Flores, M.; Isaacson, R. A.; Calvo, R.; Feher, G.; Lubitz, W. Chem. Phys. 2003, 294, 401. (41) Ishikita, H.; Knapp, E. W. J. Biol. Chem. 2003, 278, 52002. (42) Ishikita, H.; Knapp, E. W. J. Am. Chem. Soc. 2005, 127, 14714. (43) Lin, T. J.; O’Malley, P. J. J. Mol. Struct.-Theochem. 2008, 870, 31. (44) Sinnecker, S.; Flores, M.; Lubitz, W. Phys. Chem. Chem. Phys. 2006, 8, 5659. (45) Warncke, K.; Gunner, M. R.; Braun, B. S.; Gu, L. Q.; Yu, C. A.; Bruce, J. M.; Dutton, P. L. Biochemistry 1994, 33, 7830. (46) Gunner, M. R.; Madeo, J.; Zhu, Z. Y. J. Bioenerg. Biomembr. 2008, 40, 509.

Weyers et al. (47) Breton, J.; Boullais, C.; Burie, J. R.; Nabedryk, E.; Mioskowski, C. Biochemistry 1994, 33, 14378. (48) Breton, J.; Burie, J. R.; Berthomieu, C.; Berger, G.; Nabedryk, E. Biochemistry 1994, 33, 4953. (49) Deligiannakis, Y.; Hanley, J.; Rutherford, A. W. J. Am. Chem. Soc. 1999, 121, 7653. (50) Levanon, H.; Mobius, K. Annu. ReV. Biophys. Biomol. Struct. 1997, 26, 495. (51) Kolling, D. R. J.; Samoilova, R. I.; Holland, J. T.; Berry, E. A.; Dikanov, S. A.; Crofts, A. R. J. Biol. Chem. 2003, 278, 39747. (52) Lakshmi, K. V.; Brudvig, G. W. Electron Paramagnetic Resonance Distance measurements in Photosystems. In Distance Measurements in Biological Systems by EPR; Berliner, L. J., Eaton, S. S., Eaton, G. R., Eds.; Kluwer Academic/Plenum Publishers, New York, NY, 2000; Biological Magnetic Resonance Vol. 19, p 513. (53) Lakshmi, K. V.; Brudvig, G. W. Curr. Opin. Struct. Biol. 2001, 11, 523. (54) Dikanov, S. A.; Tsvetkov, Y. D. Electron Spin Echo EnVelope Modulation (ESSEM) Spectroscopy; CRC Press: Ann Arbor, MI, 1992. (55) Prisner, T.; Rohrer, M.; MacMillan, F. Annu. ReV. Phys. Chem. 2001, 52, 279. (56) Deligiannakis, Y.; Rutherford, A. W. Biochim. Biophys. ActaBioenerg. 2001, 1507, 226. (57) Connelly, N. G.; Geiger, W. E. Chem. ReV. 1996, 96, 877. (58) Bard, A. J.; Faulkner, L. R. Electrochemical Methods: Fundamentals and Applications; John Wiley & Sons, Inc.: New York, 2001. (59) Schweiger, A.; Jeschke, G. Principles of Pulse Electron Paramagnetic Resonance; Oxford University Press: New York, NY, 2001. (60) Stoll, S.; Schweiger, A. J. Magn. Reson. 2006, 178, 42. (61) Gagne, R. R.; Koval, C. A.; Lisensky, G. C. Inorg. Chem. 1980, 19, 2854. (62) Zheng, M.; Dismukes, G. C. Biochemistry 1996, 35, 8955. (63) MacMillan, F.; Lendzian, F.; Lubitz, W. Magn. Reson. Chem. 1995, 33, S81. (64) Dikanov, S. A.; Samoilova, R. I.; Kolling, D. R. J.; Holland, J. T.; Crofts, A. R. J. Biol. Chem. 2004, 279, 15814. (65) Dikanov, S. A.; Bowman, M. K. J. Magn. Reson. A 1995, 116, 125. (66) Burghaus, O.; Plato, M.; Rohrer, M.; Mobius, K.; Macmillan, F.; Lubitz, W. J. Phys. Chem. 1993, 97, 7639. (67) Wertz, J.; Bolton, J. Electron Spin Resonance; Chapman and Hall: New York, 1986. (68) McConnell, H. M.; Chesnut, D. B. J. Chem. Phys. 1958, 28, 107. (69) Das, M. R.; Connor, H. D.; Leniart, D. S.; Freed, J. H. J. Am. Chem. Soc. 1970, 92, 2258. (70) McLachlan, A. D. Mol. Phys. 1958, 1, 233. (71) Kaupp, M.; Remenyi, C.; Vaara, J.; Malkina, O. L.; Malkin, V. G. J. Am. Chem. Soc. 2002, 124, 2709.

JP907379D